3
$\begingroup$

Let $X$ be a (finite if you want) set and form the simplicial set $F^{\bullet}(X)$ with $$ F^{n}(X) = \mathrm{Hom}_{\mathrm{set}} ([n], X) $$ where the right hand side denotes arbitrary maps of sets (of course it wouldn't make sense to say order preserving as $X$ doesn't come with an order).

I'm wondering about a description of $F^{\bullet}(X)$. For example if $X = \{0,1\}$ then there are 2 0-simplices, may as well call them $[0] and [1]$ and 2 1-simplices $[0, 1]$ and $[1,0]$ glued together to form a copy of $S^1$.

Edit: as pointed out in Goodwillie's answer, this is not the end of the story, there are way more higher dimensional non-degenerate simplices.

Is there an analogous description when $X = \{0, 1, 2\}$?

A closely related question is whether there's a right adjoint to the forgetful functor from the simplex category $\Delta$ (finite ordered sets) to, say, finite (unordered) sets -- and if so what is it.

Example where such simplicial sets arise: given a map of topological spaces $f: X \to Y$ we can always form a simplicial object $\mathcal{S}^{\bullet}(f)$ with $$ \mathcal{S}^{n} = \prod\nolimits_{X}^{n} = \underbrace{X \times_{Y} \cdots \times_{Y} X}_{n\text{ times }} $$ with face and degeneracy maps given by projections and diagonals respectively. Taking connected components gives a simplicial set.

When $Y$ is the union $\bigcup_{i=1}^{N} H_{i}$ of the coordinate hyperplanes in $\mathbb{C}^{N}$ and $f: X=\coprod_{i=1}^{N} H_{i} \to \bigcup_{i=1}^{N} H_{i}=Y$ is the obvious map, I believe the simplicial set we get is $F^{\bullet}(\{1, \dots, n\})$.

$\endgroup$
8
$\begingroup$

You are overlooking some nondegenerate simplices. For example, when $X={0,1}$ there are the $2$-cells $[0,1,0]$ and $[1,0,1]$. In fact, the thing you call $F^\bullet(X)$ is infinite dimensional if $X$ has more than one element.

It is contractible whenever $X$ is non-empty; this can be seen by identifying it with the nerve of a category, a category equivalent to the point category with one morphism.

If $X=G$ has a group structure then $F^\bullet(G)$ is often called $EG$; it is a contractible space with free $G$-action.

$\endgroup$
  • 1
    $\begingroup$ Ah! Absolutely overlooked those, many thanks. Quick follow up: as an intermediate step, can we view $F^\bullet(X)$ as the nerve of the category with objects the points $x \in X$ and with a unique morphism $x \to y$ for every 2 points $x, y \in X$ (this is at least the case when $X= G$ a group and we build $EG$ as the nerve of the Cayley graph I think...)? Then including a point $x \in X$ with its identity morphism would give an equivalence of categories ... $\endgroup$ – cgodfrey Apr 29 at 2:26
  • 1
    $\begingroup$ Yes. That's what I meant in my second paragraph! $\endgroup$ – Tom Goodwillie Apr 29 at 20:55

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.